Improved hydrogen storage properties of LiAlH4 by mechanical milling with TiF3

Article abstract of DOI:10.1016/j.jallcom.2015.06.036

Dehydrogenation behavior of LiAlH₄ (lithium alanate) admixed with TiF₃ is investigated by pressure-compositionerature (PCT), fourier transform infrared spectroscopy (FTIR), X-ray Diffraction (XRD), differential scanning calorimetry (

Full text of DOI:10.1016/j.jallcom.2015.06.036

Journal of Alloys and Compounds 647 (2015) 756e762
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: http://www.elsevier.com/locate/jalcom
Improved hydrogen storage properties of LiAlH4 by mechanical
milling with TiF3
Lei Zang, Jiaxing Cai, Lipeng Zhao, Wenhong Gao, Jian Liu*, Yijing Wang
Institute of New Energy Material Chemistry, Key Laboratory of Advanced Energy Materials Chemistry (MOE), Tianjin Key Lab of Metal, Nankai University,
94 Weijin Road, Tianjin 300071, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Dehydrogenation behavior of LiAlH4 (lithium alanate) admixed with TiF3 is investigated by pressure-
composition-temperature (PCT), fourier transform infrared spectroscopy (FTIR), X-ray Diffraction
(XRD), differential scanning calorimetry (DSC) and temperature programmed desorption (TPD). The TiF3
addition enhances kinetics of LiAlH4 and decreases the decomposition temperature. The LiAlH4-2 mol %
TiF3 sample starts to release hydrogen at about 35 ꢀC and the dehydrogenation rate reaches a maximum
value at 108.4 ꢀC, compared with 145 ꢀC and 179.9 ꢀC for the as-received LiAlH4. As for the dehydro-
genation kinetics, the LiAlH4-2 mol % TiF3 sample releases about 7.0 wt % H2 at 140 ꢀC within 80 min. In
comparison, the as-received LiAlH4 sample releases only 0.8 wt % hydrogen under the same conditions.
The existence of proposed catalyst, Al3Ti formed in-situ in the process of dehydrogenation, has been
confirmed experimentally by XRD measurements. The activation energy of LiAlH4-2 mol % TiF3 com-
posite is deduced to be 66.76 kJ molꢁ1 and 88.21 kJ molꢁ1 for the first and second reaction stages of
LiAlH4 dehydrogenation.
Received 23 March 2015
Received in revised form
2 June 2015
Accepted 4 June 2015
Available online 10 June 2015
Keywords:
Hydrogen storage materials
Energy storage
Complex hydride
LiAlH4
TiF3
Ball milling
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
In recent years, light weight complex hydrides are receiving
increasing attention owing to their high gravimetric and volumetric
With the rapid global economic development, the fossil energy
is in an accelerating pace towards exhaustion. CO2 and other
harmful gases from the burning of fossil fuels, can lead to the global
warming and environmental pollution problem. In order to solve
these problems, many countries around the world are positively
developing renewable energy and clean energy. Hydrogen is an
ideal energy carrier due to its storability, environmental friendli-
ness, high calorific value and cost effectiveness [1,2]. Hydrogen
generation and storage are considered as one of the main problems
which have to be solved in the future development of hydrogen
technology [3e11]. It is crucial to find new applicable hydrogen
storage materials with high performances. Solid state hydrogen
storage materials have large volume density, good safety perfor-
mance, ease of use and transportation advantages. Therefore, future
research will focus on the safety of hydrogen storage and devel-
opment of new solid hydrogen storage materials with high-
performance [12]. According to the goal of DOE for 2015, the
target of hydrogen capacity is 6.5 wt. % [13].
hydrogen capacities [14]. Among them, LiAlH4 is regarded as one of
the most promising hydrogen storage materials owing to its high
theoretical hydrogen storage capacity (10.5 wt %). LiAlH4 material
has the following advantages: higher theoretical hydrogen storage
capacity; lower decomposition reaction enthalpy change, which
lets it release hydrogen at relatively low temperature [15]. Dehy-
drogenation of LiAlH4 occurs as given in the following reaction:
3LiAlH4/Li3AlHl6þ2Al þ 3H2 ð5:3 wt % H2Þ
Li3AlH6/3LiH þ Al þ 1:5H2 ð2:6 wt % H2Þ
LiH þ Al/LiAl þ 1:5H2 ð2:6 wt % H2Þ
R1
R2
R3
It releases 5.3 wt % H2 in the first step at 150 ꢀCe175 ꢀC, and in
the second step it releases 2.6 wt % H2 at 180 ꢀCe220 ꢀC [16e18].
The whole process of R3 only occurs in vacuum and releases 2.6 wt
% H2 at above 400 ꢀC, which is considered to be unsuited for various
applications. Thus, it is commonly accepted that only the first two
desorption processes of LiAlH4 are considered. However, there are
also many disadvantages for LiAlH4 material, such as: extremely
* Corresponding author.
E-mail address: liujian@nankai.eud.cn (J. Liu).
http://dx.doi.org/10.1016/j.jallcom.2015.06.036
0925-8388/© 2015 Elsevier B.V. All rights reserved.

L. Zang et al. / Journal of Alloys and Compounds 647 (2015) 756e762
757
high plateau pressure at relatively low temperatures, which makes
the hydride practically irreversible; the presence of an exothermic
decomposition reaction [15]. In addition, the hydrogen release rate
is relatively slow, which can be improved by ball milling and
doping. At present, a lot of research has been carried out on doping
modification. LiAlH4 starts to decompose at 61 ꢀC with 3 mol %
NiFe2O4 as an additive and releases 7.2 wt % H2 at 180 ꢀC [19].
Balema et al. mechanically ball milled LiAlH4-3 mol % TiCl4 at room
temperature for 5 min, and then LiAlH4 has completely broken
down to Li3AlH6, Al and H2 [20]. So far, the reported catalysts for
LiAlH4 are as given in the following: (1) metal hydrides, such as
MgH2 [21]; (2) elemental metals, such as Al [22], Ti [23], Fe [24], Ni
[25], and Sc [23]; (3) metal halides, such asVCl3 [26] and [27], TiF3
[28], TiCl3 [29,32] and TiCl4 [20] and [30]; (4) alloys, such as TiAl3
and Ti3Al [23]; (5) metal oxides, such as Nb2O5 and Cr2O3 [31],
Fe2O3 and Co2O3 [29,32]; and (6) others, such as TiC and TiN [33],
graphite [34], MnFe2O4 [35] and MWCNTs [36].
However, the catalytic mechanism of TiF3 on LiAlH4 is still an
unsolved problem. It is thought that TiF3 reacts with LiAlH4 to form
Al3Ti [28], and Al3Ti is considered to be an effective catalyst for the
dehydrogenation of LiAlH4 [22]. The formation of nano/micro-
crystalline Al3Ti in the milling process of 4:1 LiAlH4/TiCl4 and 3:1
NaAlH4/TiCl3 mixtures has been reported by Balema et al. [37] and
Majzoub et al. [38]. but Al3Ti cannot be detected experimentally for
samples of a typical doping level <5 mol %. To our best knowledge,
there is no report on the in-situ formation of Al3Ti during the
process of dehydrogenation. However, another mechanism of
fluoride substitution has been reported for Na3AlH6 and NaBH4
[39e41]. Fluorine partially substitutes hydrogen to form
Na3AlH6ꢁxFx and NaBH4ꢁxFx. It may lead to destabilization, which
tends to facilitate hydrogen desorption. In the present work, TiF3 (1,
2, 4, 6 mol %)-LiAlH4 have been ball milled under certain conditions.
Samples have been analyzed by Pressureecompositione temper-
ature (PCT) apparatus, differential scanning calorimetry (DSC), X-
ray diffraction (XRD), fourier transform infrared spectroscopy
(FTIR) and temperature programmed desorption (TPD). The kinetic
and thermodynamic performances of the LiAlH4eTiF3 composite
will be further explored, and the in-situ formation of Al3Ti in
dehydrogenation will be discussed.
Differential scanning calorimetry (DSC) was performed using a TA
apparatus (DSC Q20P) in a flow (50 ml minꢁ1) of high purity Ar,
with about 6 mg sample for each measurement.
3. Results and discussion
3.1. Dehydrogenation properties
3.1.1. Effect of the doping ratios on dehydrogenation
Firstly, the LiAlH4-xTiF3 composites (x ¼ 0, 1, 2, 4, 6 mol %,
respectively) have been prepared to optimize the addition amount
of TiF3, with x standing for the mole percentage of TiF3 relative to
LiAlH4. For comparison, the dehydrogenation profiles of the doped
samples and as-received LiAlH4 are all presented in Fig. 1(A).
Obviously, TiF3 enormously enhances the dehydrogenation kinetics
of commercial LiAlH4. The dehydrogenation rate becomes faster
with the adding amount of TiF3 increasing. As shown from profile
‘a’ in Fig. 1(A), dehydrogenation of the undoped-LiAlH4 sample
occurs obviously in two steps (R1 and R2) and releases 7.4 wt % H2
within 30 min at 250 ꢀC. For comparison, LiAlH4-1 mol % TiF3 re-
leases 6.96 wt % H2 within 20 min; LiAlH4-2 mol % TiF3 releases
7.25 wt % H2 within 18 min; LiAlH4-4 mol% TiF3 releases 6.82 wt %
H2 within 13 min; LiAlH4-6 mol % TiF3 releases 6.16 wt % H2 within
10 min. As we can see, with increasing addition amount of TiF3,
dehydrogenation rate of LiAlH4 increases considerably.
In contrast to the dehydrogenation rate, the H2 release amount
of LiAlH4 decreases and is lower than that of the as-received LiAlH4.
We think that TiF3 is a beneficial catalyst and effectively improves
the dehydrogenation kinetics of pure LiAlH4. As seen in Fig. 1(B),
our initial speculation has been verified that more addition of TiF3
further reduces the amount of released hydrogen when the H2
capacity is calculated on the basis of the total mass of samples.
Taking the dehydrogenation kinetics and hydrogen capacity into
consideration, the LiAlH4-2 mol % TiF3 composite is the best
compromise to be further explored.
Fig. 2 shows FTIR spectra of the 0.5 h ball-milled LiAlH4 doped
with 0, 1, 2, 4, 6 mol % TiF3. For the undoped-LiAlH4, the peaks at
883 and 705 cmꢁ1 are the bending modes of [AlH4]ꢁ, and the peaks
at 1780 and 1621 cmꢁ1 are the stretching modes of [AlH4]ꢁ. There is
no vibration peak of Li3AlH6. Thus, we can consider that as-milled
LiAlH4 is stable and no Li3AlH6 formed during ball milling. In
comparison with the asemilled LiAlH4, it is obvious that the [AlH4]ꢁ
bands for the LiAlH4eTiF3 composite are still in the same positions,
however the peak intensities decrease. Obviously, there are IR vi-
bration peaks (v3 [AlH6]3ꢁ) at 1402 and 1298 cmꢁ1 which are the
stretching modes of [AlH6]3ꢁ, suggesting that a certain amount of
LiAlH4 decomposes into Li3AlH6, resulting from a small amount of
hydrogen release during ball-milling. This is in accordance with the
results of PCT that the hydrogen capacity of LiAlH4eTiF3 composite
is lower than the as-received LiAlH4 after ballemilling. As we can
see, the stretching modes of [AlH6]3ꢁ are weak, indicating a small
amount of Li3AlH6. That is to say, only a small part of LiAlH4 de-
composes during high-energy ball milling process. Obviously, the
peak intensity of the [AlH6]3ꢁ increases with the content of TiF3,
indicating that the decomposition amount of LiAlH4 increases. The
addition of TiF3 benefits reaction (R1) for LiAlH4 during the ball
milling.
2. Experimental
The starting materials, LiAlH4 (95%) were purchased from Acros
Corp. and TiF3 (99%) were purchased from Alfa Aesar. All the ma-
terials were used without further purification. LiAlH4 was me-
chanically milled with x mol % (x ¼ 1, 2, 4, 6) TiF3 for 0.5 h under
1 MPa H2 atmosphere using a QM-3SP2 planetary ball mill at
300 rpm in a stainless steel vessel. The ball-to-powder ratio was
60:1. For comparison, pure LiAlH4 was also mechanically milled
under identical conditions. All sample operations were performed
in a glove box under Ar (99.999% purity) and the H2O and O2
contents were kept below 1 ppm.
XRD analysis was carried out using a powder X-ray Diffraction
(XRD, Rigaku D/Max PC2500, Cu Ka radiation, 40 kV, 100 mA) at a
scanning rate of 4ꢀ per min over the range of 20e80ꢀ2
q. The sample
holder was covered with Mylar film to protect the sample from
moisture and atmospheric oxygen during transportation. Temper-
ature Programmed Desorption (TPD) of H2 was performed using a
home-made apparatus, which measures gas release rate at various
temperatures. About 70 ꢁm1 g sample was used and heated at a
ramping rate of 2 ꢀC min in a 35 mL/min Ar flow while heating
from 30 to 260 ꢀC. Dehydrogenation kinetics was measured by
pressure-composition-temperature (PCT) apparatus at certain
temperatures. FTIR spectra were measured by using an FTIR-650
3.1.2. Effect of placing for different time and temperature
As reported in literature, the initial temperature of LiAlH4-4 mol
% TiF3 composites after ball milling is 80 ꢀC [28]. However a special
phenomenon was found in our experimental process, the decom-
position amount of LiAlH4 obviously decreased, after the ball-
milled materials were placed at room temperature for a period of
time. As shown in Fig. 3(A), if we make the dehydrogenation test of
spectrometer (Tianjin Gangdong) at a resolution of 4 cmꢁ1
.

758
L. Zang et al. / Journal of Alloys and Compounds 647 (2015) 756e762
Fig. 1. The dehydrogenation profiles of the ball-milled LiAlH4-xTiF3 composites by PCT at 250 ꢀC: (a) undoped-LiAlH4; (b) LiAlH4-1 mol % TiF3; (c) LiAlH4-2 mol % TiF3; (d) LiAlH4-
4 mol % TiF3; (e) LiAlH4-6 mol % TiF3. The hydrogen capacity was calculated: (A) only on the basis of LiAlH4 and (B) after taking the weight of TiF3 into account.
the LiAlH4-2 mol % TiF3 composites at 250 ꢀC immediately after ball
milling, it releases 7.25 wt % H2; however, the amount of H2 release
decreases to 7.04 wt % after the ball-milled sample is placed in the
glove box for 5 h. This suggests that the sample probably releases
some hydrogen when placed at room temperature. We then make
the dehydrogenation test of samples which are placed in the glove
box at room temperature (about 30 ꢀC) for 10 h, 24 h, 48 h, and
120 h; and the amount of H2 release is 6.32 wt %, 4.7 wt %, 3.87 wt %
and 3.5 wt % respectively.
It would be reasonable that low temperature is beneficial to
keep the samples from decomposition. As seen from profile ‘a’ in
Fig. 4, the amount of hydrogen release from the sample kept at 0 ꢀC
for 48 h is 7.2 wt %, and it is almost the same as that from the as-
prepared sample. For comparison, the amount of hydrogen
release from the sample kept at room temperature (about 30 ꢀC) for
48 h decreases to 4.3 wt %. We think that the LiAlH4-2 mol % TiF3
composites release hydrogen considerably at room temperature.
With the addition of TiF3 catalyst, the initial dehydrogenation
temperature of LiAlH4 reduces to nearly room temperature. The
XRD tests have provided evidence for this explanation. As shown in
Fig. 3 (B), the peak at 2
diffraction intensity of Li3AlH6 is stronger with increasing time of
placing at room temperature; meanwhile, the peak intensity of Al is
q
¼ 31.6 belongs to the Li3AlH6. The
Fig. 2. FTIR spectra of the as-milled LiAlH4; LiAlH4-1 mol % TiF3, LiAlH4-2 mol % TiF3,
LiAlH4-4 mol % TiF3, LiAlH4-6 mol % TiF3 samples after ball milling.
Fig. 3. (A) The dehydrogenation profiles and (B) The XRD profiles of the ball-milled LiAlH4-2 mol % TiF3 composites placed in the glove box for different lengths of time at room
temperature.

L. Zang et al. / Journal of Alloys and Compounds 647 (2015) 756e762
759
LiAlH4-6 mol % TiF3. The dehydrogenation capacity decreases with
increasing addition amount of TiF3 from 1 to 2 mol %, and then it
decreases from 2 to 6 mol %. It reveals that the hydrogen capacity
reduces obviously with more addition of TiF3. TiF3 catalyst de-
creases the dehydrogenation temperature, with the cost of loss of
hydrogen release content. The excessive amount of TiF3 catalyst is
detrimental for hydrogen capacity. The hydrogen desorption ca-
pacity of LiAlH4-2 mol %TiF3 sample is 7.1 wt %, which is more than
those of any other doping samples. The doping of TiF3 is beneficial
for complete decomposition of LiAlH4 which leads to an increase of
hydrogen capacity. On the other hand, the loss of hydrogen prob-
ably increases with more TiF3 during ball milling, which then leads
to a decrease of hydrogen capacity. As a result of the two factors, the
hydrogen capacity of LiAlH4-2 mol % TiF3 composite reaches a
maximum. However, the hydrogen capacity of doped-LiAlH4 sam-
ples is always lower than the theoretical value (7.9 wt %). It is
because as-received LiAlH4 sample containing impurities, which
results in the hydrogen capacity not reaching the theoretical value.
On the other hand, the TiF3 also makes LiAlH4 decompose to Li3AlH6
and release a small amount of hydrogen during ball-milling, which
has been confirmed by the IR spectra. Meanwhile, R1 becomes
slower and goes on over a wider temperature range and the
boundaries of R1 and R2 becomes more and more undefined, it is
probably because R2 already starts before R1 finishes completely
[28]. The details of TPD dehydrogenation curves are shown in
Table 1.
Fig. 4. The dehydrogenation profiles at 250 ꢀC of the ball-milled LiAlH4-2 mol % TiF3
composites (a) kept with ice bath at 0 ꢀC; (b) placed in the glove box at room tem-
perature for 48 h, respectively.
stronger and the peak of LiAlH4 decreases. The first reaction (R1) of
LiAlH4-2 mol % TiF3 composites takes place at room temperature.
3.1.3. TPD measurements
TPD tests further prove our speculation. Fig. 5 shows TPD curves
of the ball-milled LiAlH4 doped with 0, 1, 2, 4, 6 mol % TiF3. As
shown in Fig. 5(A), there are two distinct hydrogen desorption
peaks for five samples when heating from 30 ꢀC to 260 ꢀC, which is
in accordance with the two-step dehydrogenation reaction (R1 and
R2) of LiAlH4. The onset temperature of the undoped-LiAlH4 sample
is around 145 ꢀC with the maximum at 179.9 ꢀC. The LiAlH4-2 mol %
TiF3 sample starts to release hydrogen at around 35 ꢀC with the
maximum at 108.4 ꢀC, which is 71.5 ꢀC lower than that for the
undoped-LiAlH4 sample (179.9 ꢀC). It has proved that the onset
temperature of LiAlH4-2 mol % TiF3 composites has reduced to
nearly room temperature. Furthermore, the peak of dehydrogena-
tion shifts to lower temperatures with TiF3 amount increasing. The
peak of the first reaction for LiAlH4eTiF3 is a broad peak. For the
second dehydrogenation reaction, the signal intensities become
weaker with increasing TiF3 amount, and the dehydrogenation
temperature just decreases a little. It is clear that TiF3 is an effective
catalyst for the decrease of desorption temperature.
3.2. Dehydrogenation kinetics and thermodynamics
The dehydrogenation kinetics of the LiAlH4-2 mol % TiF3 sample
under relatively lower temperatures are shown in Fig. 6, while the
kinetic behavior of undoped-LiAlH4 is also included. The results
show that the LiAlH4-2 mol % TiF3 composite releases 1.75, 3.25, 4.1,
5.1, 6.77 and 7.0 wt % H2 at 45, 60, 80, 100, 120 and 140 ꢀC,
respectively. The hydrogen capacity becomes higher with higher
desorption temperatures. Acting as a catalyst, TiF3 enormously
decreases the dehydrogenation temperature. As can be seen from
Fig. 6, at 140 ꢀC, the LiAlH4-2 mol % TiF3 sample releases 7 wt % H2
within 80 min, whereas the undoped LiAlH4 sample (without ball
milling) only releases 0.8 wt % H2 within the same time. The
hydrogen capacity of the undoped LiAlH4 is only 4.4 wt % at this
temperature. At 45 ꢀC, the LiAlH4-2 mol % TiF3 sample releases
about 1.75 wt % H2 within 8 h; on the other hand, the undoped
sample cannot release H2 at this temperature. The above results
show that the dehydrogenation rate of LiAlH4-2 mol % TiF3 is much
faster than that of the undoped LiAlH4. The improved kinetics is
Fig. 5(B) presents the TPD dehydrogenation capacity of LiAlH4,
LiAlH4-1 mol % TiF3, LiAlH4-2 mol % TiF3, LiAlH4-4 mol % TiF3 and
Fig. 5. TPD dehydrogenation curves of undoped-LiAlH4, LiAlH4-1 mol % TiF3, LiAlH4 e 2 mol % TiF3, LiAlH4-4 mol % TiF3, LiAlH4-6 mol % TiF3 (Heating ramp 2 ꢀC/min).

760
L. Zang et al. / Journal of Alloys and Compounds 647 (2015) 756e762
Table 1
Details of TPD dehydrogenation curves of undoped-LiAlH4, LiAlH4-1 mol % TiF3, LiAlH4-2 mol % TiF3, LiAlH4-4 mol % TiF3, LiAlH4-6 mol % TiF3.
Initial temp. (ꢀC)
Dehydrogenation peak of R1 (ꢀC)
Dehydrogenation peak of R2 (ꢀC)
Total amount of H2 release (%)
LiAlH4
145
35
35
35
35
179.9
114.6
108.4
88.1
216.6
202.2
201.9
198
7.3
6.8
7.1
6.7
5.9
LiAlH4-1 mol % TiF3
LiAlH4-2 mol % TiF3
LiAlH4-4 mol % TiF3
LiAlH4-6 mol % TiF3
80.5
194
has been determined with the Kissinger's approach [16]. Fig. 7(B)
shows the Arrhenius plots for the two endothermic peaks of LiAlH4-
2 mol % TiF3 sample. Ea of the two dehydrogenation reactions for
LiAlH4 is deduced to be 66.76 kJ molꢁ1 and 88.21 kJ molꢁ1, which
are 49.44 kJ molꢁ1 and 44.79 kJ molꢁ1 lower than the theoretical
value of pure LiAlH4 (116.2 kJ molꢁ1 and 133.0 kJ molꢁ1) [35]. This
reveals that the TiF3 enormously reduces the activation energy of
LiAlH4 and obviously improves the dehydrogenation kinetics of
LiAlH4-2 mol % TiF3 composite.
3.3. Mechanism analysis
Fig. 8 shows the XRD patterns of as-received LiAlH4, as-milled
LiAlH4, as-milled LiAlH4-2 mol % TiF3 and LiAlH4-2 mol % TiF3 af-
ter dehydrogenation. The small peak at 2
q
¼ 34.8 is the impurity
LiCl, which could explain the results of PCT and TPD that the
dehydrogenation capacity of doping LiAlH4 is lower than theoret-
ical capacity. As seen from profile ‘b’ in Fig. 8, the peak positions of
as-milled LiAlH4 were the same as those of the as-received LiAlH4
except for the decreasing of intensity. That is to say, ball-milling
does not make LiAlH4 decompose and just make the particle size
Fig. 6. The PCT dehydrogenation characteristics of undoped LiAlH4 sample at 140 ꢀC
and ball-milled LiAlH4-2 mol % TiF3 sample at various temperatures.
probably owing to the defect creation, formation of some kind of
active material, and particle size reduction during milling
[28,42e44]. Moreover, the catalytic properties of TiF3 is superior to
some previously documented catalysts, such as TiCl4 [20],
elemental metals [22e25], carbon material [34,36].
To further investigate dehydrogenation kinetics and thermo-
dynamics of LiAlH4-2 mol % TiF3 sample, the DSC curves were
measured with heating from 50 ꢀC to 260 ꢀC at various rates of 2, 5,
8 ꢀC/min, respectively. In the aspect of thermodynamics, as shown
in Fig. 7 (A), DSC curves of LiAlH4-2 mol % TiF3 sample exhibit two
distinct endothermic peaks at heating ramp of 2 ꢀC/min. Clearly, the
first peak of LiAlH4-2 mol % TiF3 sample is at about 125.81 ꢀC and
the second is at 179.83 ꢀC, which corresponds to the two-step
decomposition reactions. To explore how the TiF3 catalyst acts on
the kinetic performance of LiAlH4, Ea of LiAlH4-2 mol % TiF3 sample
smaller. As seen from profile ‘c’, the small peak at 2
q
¼ 31.6 belongs
to the Li3AlH6. A small amount of LiAlH4 has already decomposed to
Li3AlH6, which is consistent with the result of FTIR. However, it
should be noted that no F-containing and Ti-containing phases are
detected after ball milling or after dehydrogenation. It is probably
because that their amount is beyond the detection limit of XRD
technique or it is in amorphous state. The following reactions
probably take place, according to the mechanism proposed by S.-S.
Liu et al. [28].
3LiAlH4 þ TiF3/3LiF þ Al3Ti þ 6H2
3LiAlH4 þ TiF3/3LiF þ 3Al þ TiH2 þ 5H2
3LiAlH4 þ TiF3/3LiF þ 3Al þ Ti þ 6H2
R4
R5
R6
Fig. 7. (A) DSC profiles of LiAlH4-2 mol % TiF3 sample(2, 5 and 8 ꢀC minꢁ1); (B) Kissinger plots for the dehydrogenation of LiAlH4-2 mol % TiF3 sample (A: LiAlH4/Li3AlH6；B:
Li3AlH6/LiH).

L. Zang et al. / Journal of Alloys and Compounds 647 (2015) 756e762
761
milling [37,38], the Al3Ti formed in-situ during dehydrogenation
process has been detected by XRD for the first time to our best
knowledge. Thus, we can believe that LiAlH4 reacts with TiF3 as
reaction R4, not R5 or R6, and Al3Ti forms in the process of dehy-
drogenation when the ratio is 3:1 at certain temperatures; and the
in-situ formed Al3Ti catalyzes dehydrogenation through reaction
R1 and R2. However, the dehydrogenation capacity cannot reach
the theoretical value (10.5 wt %), it is probably because the as-
received LiAlH4 is not pure and some LiAlH4 still decomposes as
R1 and R2, which lowers the dehydrogenation capacity of the
samples.
For comparison, the ball-milled sample releases only 2.2 wt % H2
at 250 ꢀC, and Al3Ti cannot be found in the XRD patterns both
before and after dehydrogenation. However, there was a small
amount of LiAlH4 remaining after ball milling. We can deduce that
only a small amount of LiAlH4 reacts with TiF3 as R4 during ball
milling even though the ratio is 3:1. It is probably because the slow
speed and the small energy of ball milling could not make the re-
action complete. And then the formed Al3Ti catalyzes the LiAlH4
decompose, whereas the amount of Al3Ti is still beyond the
detection limit of XRD, or it is amorphous in the samples. From the
results above, we can verify the formation of crystalline Al3Ti in
reaction R4.
Fig. 8. XRD patterns of the samples (a) as-received LiAlH4; (b) as-milled LiAlH4; (c) as-
milled LiAlH4-2 mol % TiF3; (d) LiAlH4-2 mol % TiF3 after dehydrogenation.
Kang et al. conclude that the formation of Al3Ti is favorable.
4. Conclusions
Because Al3Ti
(D ) is more stable than TiH2
Gf.136 kJ molꢁ1
(D
Gf.86 kJ molꢁ1) [45]. It is thought that only some of TiF3 reacted
It has been demonstrated that TiF3 obviously enhances the
desorption kinetics of as-received LiAlH4. The addition of 2 mol %
TiF3 leads to the onset temperature of LiAlH4 decreasing to around
35 ꢀC and the dehydrogenation rate reaches a maximum value at
108.4 ꢀC. As for the dehydrogenation kinetics, the LiAlH4-2 mol %
TiF3 sample releases about 7.0 wt % H2 at 140 ꢀC within 80 min. The
with LiAlH4 to form the catalyst Al3Ti during ball-milling and Al3Ti
catalyzed the LiAlH4 decomposition to Li3AlH6. However, Al3Ti is
not detected after dehydrogenation. It is probably because the
diffractions of Al3Ti are weak due to its large peak widths, high
dispersion, some amorphous state or small amount [46]. However,
another view indicates that fluorine substitute hydrogen to form
the stable structures of AlF6ꢁ and AlFꢁ4 , which tends to facilitate
hydrogen desorption [39e41].
To further investigate catalytic mechanism of TiF3, the PCT and
XRD curves were measured with the LiAlH4eTiF3 samples of
3:1 mol ratio. As seen from profile ‘b’ in Fig. 9(A), the 3LiAlH4eTiF3
sample with physical mixture quickly releases nearly 9 wt % H2
(larger than 7.9 wt %, which is the sum of hydrogen release from R1
and R2) within 2 min at 250 ꢀC. Correspondingly, the diffraction
peaks of Al3Ti and LiF are present in profile ‘d’ in Fig. 9(B). Although
the formation of Al3Ti has been confirmed during the process of ball
Ea values of the LiAlH4-2 mol
% TiF3 sample decrease to
66.76 kJ molꢁ1 and 88.21 kJmolꢁ1 for the two hydrogen desorption
reaction of LiAlH4, resulting in a faster dehydrogenation rate. The
presence of Li3AlH6 is confirmed by FTIR and XRD after ball milling,
demonstrating that the addition of TiF3 leads to the decomposition
of LiAlH4 during ball milling. On reaction mechanism, the in-situ
formed Al3Ti has been found in XRD patterns of the mixed
3LiAlH4eTiF3 samples without ball milling, after dehydrogenation
at 250 ꢀC. TiF3 presumably reacts with LiAlH4 to form the Al3Ti and
LiF, following the proposed reaction R4, and then the formed Al3Ti
catalyzes LiAlH4 to decompose.
Fig. 9. (A) PCT profiles of 3LiAlH4eTiF3 sample at 250 ꢀC (a: after ball milling; b: physical mixture without ball milling); (B) XRD profiles of 3LiAlH4eTiF3 sample (a: the ball-milled
sample before dehydrogenation, b: the ball-milled sample after dehydrogenation; c: the physically-mixed sample before dehydrogenation, d: the physically-mixed sample after
dehydrogenation).

762
L. Zang et al. / Journal of Alloys and Compounds 647 (2015) 756e762
[22] M. Resan, M.D. Hampton, J.K. Lomness, D.K. Slattery, Int. J. Hydrogen Energy
Acknowledgment
30 (2005) 1417e1421.
[23] H.W. Langmi, G.S. McGrady, X.F. Liu, C.M. Jensen, J. Phys. Chem. C 114 (2010)
10666e10669.
The authors thank the Tianjin Natural Science Foundation
[24] A.V. Robert, R. Parviz, Int. J. Hydrogen Energy 37 (2012) 9088e9102.
[25] X.P. Zheng, X.H. Qu, I.S. Humail, P. Li, G.Q. Wang, Int. J. Hydrogen Energy 32
(2007) 1141e1144.
09JCZDJC24800 for the financial support for our research.
[26] D. Blanchard, H.W. Brinks, B.C. Hauback, P. Norby, Mater. Sci. Eng. B 108
(2004) 54e59.
References
[27] J.R. Ares, K.F. Aguey-Zinsou, M. Elsaesser, X.Z. Ma, M. Dornheim, T. Klassen, Int.
J. Hydrogen Energy 32 (2007) 1033e1040.
[28] S.S. Liu, L.X. Sun, Y. Zhang, F. Xu, J. Zhang, H.L. Chu, M.Q. Fan, T. Zhang,
X.Y. Song, J.P. Grolier, Int. J. Hydrogen Energy 34 (2009) 8079e8085.
[29] B.A. Placidus, T.G. John, J.S. Patrick, A.V. Andrey, S.F. Timothy, J. Phys. Chem. C
116 (2012) 21886e21894.
[1] M.Q. Fan, S. Liu, L.X. Sun, F. Xu, S. Wang, J. Zhang, D.S. Mei, F.L. Huang,
Q.M. Zhang, Int. J. Hydrogen Energy 37 (2012) 4571e4579.
[2] H.B. Dai, G.L. Ma, H.J. Xia, P. Wang, Energy & Environ. Sci. 4 (2011) 2206e2212.
[3] K.A. Brown, S. Dayal, X. Ai, G. Rumbles, P.W. King, J. Am. Chem. Soc. 28 (2010)
9672e9680.
[4] K.M. Fair, X.Y. Cui, L. Li, C.C. Shieh, R.K. Zheng, Z.W. Liu, B. Delley, M.J. Ford,
S.P. Ringer, C. Stampfl, Phys. Rev. B 87 (2013) 014102.
[30] J. Fu, L. Rontzsch, T. Schmidt, M. Tegel, T. Weißgarber, B. Kieback, Int. J.
Hydrogen Energy 37 (2012) 13387e13392.
[5] J. Hensel, G. Wang, Y. Li, J.Z. Zhang, Nano Lett. 2 (2010) 478e483.
[6] D. Streich, Y. Astuti, M. Orlandi, L. Schwartz, R. Lomoth, L. Hammarstrom,
S. Ott, Chem. A Euro. J. 1 (2010) 60e63.
[7] Z.G. Huang, T. Autrey, Energy Environ. Sci. 5 (2012) 9257e9268.
[8] R. Khnayzer, L.B. Thompson, M. Zamkov, S. Ardo, G. Meyer, C. Murphy,
F.N. Castellano, J. Phys. Chem. C 1 (2012) 1429e1438.
[9] H.M. Chen, C.K. Chen, M.L. Tseng, P.C. Wu, C.M. Chang, L.C. Cheng, H.W. Huang,
T.S. Chan, D.W. Huang, R.-S. Liu, Small 17 (2013) 2926e2936.
[10] D.J. Bull, E. Weidner, I.L. Shabalin, M.T.F. Telling, C.M. Jewell, D.H. Gregory,
D.K. Ross, Phys. Chem. Chem. Phys. 12 (2010) 2089e2097.
[11] H. Lv, J. Song, H. Zhu, Y.V. Geletii, J. Bacsa, C. Zhao, T. Lian, D.G. Musaev,
C.L. Hill, J. Catal. 307 (2013) 48e54.
[12] B. Johnston, M.C. Mayo, A. Khare, Technovation 25 (2005) 569e585.
[13] S. Satyapal, J. Petrovic, C. Read, G. Thomas, G. Ordaz, Catal. Today 120 (2007)
246e256.
[31] Rafiuddin, X.H. Qu, P. Li, L. Zhang, A. Mashkoor, J. Phys. Chem. C 26 (2011)
13088e13099.
[32] Z.L. Li, P. Li, Q. Wan, F.Q. Zhai, Z.W. Liu, K.F. Zhao, L. Wang, S.Y. Lu, L. Zou,
X.H. Qu, A. Alex, M. Volinsky, J. Phys. Chem. C 117 (2013) 18343e18352.
[33] R.A. Varin, R. Parviz, Int. J. Hydrogen Energy 39 (2014) 2575e2586.
[34] M. Ismail, Y.X. Zhao, B. Yu, S.X. Dou, Int. J. Electroact. Mater. 1 (2013) 13e22.
[35] Q. Wan, P. Li, Z.L. Li, F.Q. Zhai, X.H. Qu, A. Alex, J. Phys. Chem. C 117 (2013)
26940e26947.
[36] W.C. Hsu, C.H. Yang, W.T. Tsai, Int. J. Hydrogen Energy 39 (2014) 927e933.
[37] V.P. Balema, J.W. Wiencha, K.W. Dennisa, M. Pruskia, V.K. Pecharsky, J. Alloy
Compd. 329 (2001) 108e114.
[38] E.H. Majzoub, K.J. Gross, J. Alloy Compd. 356 (2003) 363e367.
[39] N. Eigen, U. Boesenberg, J.B. von Colbe, J. Alloy Compd. 477 (2009) 76e80.
[40] L.H. Rude, U. Filsø, V. D'Anna, A. Spyratou, B. Richter, S. Hino, Phys. Chem.
Chem. Phys. 15 (2013) 18185e18194.
[14] L. Li, Y. Wang, F.Y. Qiu, Y.J. Wang, Y.N. Xu, C.H. An, L.F. Jiao, H.T. Yuan, J. Alloy
Compd. 566 (2013) 137e141.
[41] H.W. Brinks, A. Fossdal, B.C. Hauback, J. Phys. Chem.
C 112 (2008)
pp.5658epp.5661.
[15] R.A. Varin, L. Zbroniec, J. Alloy Compd. 504 (2010) 89e101.
[16] A. Andreasena, T. Veggea, A.S. Pedersena, J. Phys. Chem.
[42] J.R. Ares, K.F. AgueyeZinsou, M. Porcu, J.M. Sykes, M. Dornheim, T. Klassen,
Mater. Res. Bull. 43 (2008) 1263e1275.
C 116 (2012)
22327e22335.
[43] L.H. Kumar, B. Viswanathan, S.S. Murthy, Int. J. Hydrogen Energy 33 (2008)
366e373.
[44] L. Li, Y.N. Xu, Y. Wang, Y.J. Wang, F.Y. Qiu, C.H. An, L.F. Jiao, H.T. Yuan, Dalton
Trans. 43 (2014) 1806e1813.
[45] X.D. Kang, P. Wang, X.P. Song, X.D. Yao, G.Q. Lu, H.M. Cheng, J. Alloy. Compd.
424 (2006) 365e369.
[17] F.Q. Zhai, P. Li, A.Z. Sun, S. Wu, Q. Wan, W.N. Zhang, Y.L. Li, L.Q. Cui, X.H. Qu,
J. Phys. Chem. C 116 (2012) 11939e11945.
[18] J. Chen, N. Kuriyama, J. Phys. Chem. B 105 (2001) 11214e11220.
[19] P. Li, Z.L. Li, F.Q. Zhai, Q. Wan, X.Q. Li, X.H. Qu, A.V. Alex, J. Phys. Chem. C 117
(2013) 25917e25925.
[20] V.P. Balema, J.W. Wiench, K.W. Dennis, J. Alloy Compd. 329 (2001) 108.
[21] T. Czujko, Z. Zaranski, I.E. Malka, Z. Wronski, J. Alloys Compd. 509 (2011)
604e609.
[46] S.S. Liu, Y. Zhang, L.X. Sun, J. Zhang, J.N. Zhao, F. Xu, F.L. Huang, Int. J. Hydrogen
Energy 35 (2010) 4554e4561.